Award Holder: Nikolas Gestrich
University: Institute of Archaeology, University College London
Title of Research: Tongo Maare Diabal: the material culture and built space of a complex, multi-ethnic community in the Gourma (Mali; AD 400-1100).

This is the report of a season of archaeological excavation and survey near the town of Douentza, Mali, in the north of the so-called “Dogon country”. The research was aimed at producing a better understanding of the structure of societies living in this region at the beginning of the second millennium AD, and of the region's historical development more generally.

History of the Douentza region and the Dogon country:
Oral histories collected in the Dogon country since the early 20th century have given us a version of history that describes several different human groups that settled the region over time. The large sandstone massifs and inselbergs with their steep escarpments and deep fissures made this area a haven for those who were looking to escape from persecution. The Dogon report that they sought refuge here some 500 years ago, after leaving their homeland in the Mandé area when faced with the choice of converting to Islam or leaving.

On their arrival in the region of the Bandiagara plateau and the south-western Gourma they are said to have replaced a population named Tellem, whose spectacular cliff dwellings have been dated from the 11th century AD onwards. Before the Tellem, only remains from the third – second centuries BC and much older Late Stone Age remains had been found in the area.

Over the last decade, this picture has begun to change. Archaeological research in the southern part of the Dogon country and the Gondo plain has shown that, in various guises, this area was occupied during most parts of the last five thousand years. The site of Tongo Maaré Diabal on which this research focussed, was first excavated in the early 1990s and was shown to have been occupied from the 4th to the 11th centuries AD, thus immediately preceding the Tellem remains. By and by, the disciplines of historical linguistics, archaeology and anthropology are coming to realise that not only does the clear-cut version of history obtained through the Dogon not match well with their findings, but also that the Dogon as a population group are very diverse and that some of the cultural and linguistic elements found here appear to have been in place for up to three thousand years.

At Tongo Maaré Diabal, the earlier excavations had suggested the existence of two distinct cultural traditions at the site. In one of these, the techniques employed in producing and decorating pottery shows strong links with the practices of the Tellem and Dogon, while the other is closer to contemporary traditions of the Lakes Region of Mali, where urban settlements and state-level organisation mark the second half of the first millennium AD. Because the excavation campaigns of the 1990s were focused on the sequence of settlement at the site, they could not uncover much of the context within which it occurred. The aim of this fieldwork season was therefore to discover the circumstances of the occupation at Tongo Maaré Diabal: What is the relationship between the two pottery traditions? Do they correspond to separate population groups living side by side? How does the site fit into a wider archaeological landscape? What can the site and its surroundings tell us about the historical development of the area by its links to Tellem and Dogon occupations?

In the excavation, I was accompanied by a Malian co-researcher, Daouda Keita, and a collaborator, Soumaila Coulibaly, who was later replaced with Yamoussa Fané. Also accompanying us were the site foreman, who I had worked with in earlier excavations, Abdoulaye “Baba” Coulibaly, and a student from the University of Botswana on work experience, Malebogo Seone.

We arrived in Douentza, the town on whose outskirts the site lies, on the 20th of January 2010, having been delayed by bureaucracy in the capital. Much of the provisions for the project, such as the staple foods of rice, couscous, potatoes, peanut paste etc., had to be bought in Bamako, as they are either unavailable or very expensive in Douentza. We hired a house at the edge of town which would serve as accommodation, a base of operations, for storing the recovered materials and conducting their recording and primary analyses. As is usual in the area, the house had no electricity or running water, but was close to a well. The compound had a high wall, and the doors could be locked – an important consideration not so much for our belongings since theft is very rare in Mali, but rather in light of increasing reports of kidnapping of foreigners by Islamist groups in northern Mali.

With the help of pre-existing contacts, we were able to raise a workforce of 5 local men, one of which had already participated in the excavations 15 years ago. We also needed to announce our presence and intentions to the local dignitaries, both political and traditional. Thankfully, the prefect of the region, the mayor of Douentza and the traditional chief of the Dogon community were pleased with our project and offered their support.  We were able to begin excavations on the 22nd of January, and began by looking for the trenches from the previous excavations. The main exposure was easy to find, and marked out our first trench next to it. Over the next weeks this trench revealed 12 mud-brick structures, that appeared to belong to 2 or more domestic compounds and were divided by a passageway. Some of the areas appeared to have been heavily burnt, and large numbers of intact pots were left on the floors of the buildings when they were abandoned. At first, this led us to believe that the site might have experienced a violent abandonment, having possibly been burnt down by raiding parties. This was the interpretation reached after the first excavations at the site, and is known from contemporary sites in the wider region.

Since we wanted to excavate as large a contiguous area as possible, preferably an entire compound, we extended our trench to the Northwest. This area revealed what Yamoussa Fané, himself a trained blacksmith, was instantly able to interpret as an ancient blacksmith's workshop. Our continuing excavations confirmed this diagnosis: we found four hearths or furnace structures, a clay bellow emplacement of a type still in use in modern-day Mali, large amounts of charcoal and slag, and evidence on the floors and the walls of burning at very high temperatures. In 1996, an iron smelting furnace had been excavated only two kilometres from the site. Our discovery now meant that we would have the opportunity to study an entire iron production sequence from the smelting through the smithing to the finished object, and gain a unique insight into a technology of the highest importance in pre-colonial West Africa.

Also, our interpretation of violent abandonment had to be re-evaluated in the light of this evidence. After the find of the blacksmith's workshop, it seemed to us that the many burning incidents visible at the site's surface were in fact related to metallurgy rather than the burning of dwellings. Closer inspection revealed the unique dimpled bricks used in the excavated forge structures, and the remains of tuyeres (air pipes for inducing draught into the fire) wherever we found burnt structures eroding on the site's surface. It seemed as though the inhabitants of Tongo Maaré Diabal were heavily involved in ironworking, and that there was far more than the usual one or two village smiths present at the site.

We went on to open a further exposure on the Western end of the site, such as to form a point of comparison with the data recovered from the main excavation area. Much of the final argument is expected to rest on the main exposure, where the individual living spaces will be separated to give an impression of the organisation and use of space in architecture and artefacts. By this method, we hope to be able to gain an insight into the social organisation of the inhabitants of Tongo Maaré Diabal, and into the circumstances of their life at the site, when the analyses and interpretation of all the materials recovered are complete. However, this area of occupation and the finds made within it need to be contextualised within the occupation of the site itself, as they might not be representative of the situation in the rest of the settlement; hence the excavation of the second unit. Similarly, the site itself may not be representative of the wider area, which is why the excavations were complimented by a program of survey in a five kilometre radius around the site.

At the second exposure, we uncovered four buildings and what appeared to be one outdoor area. As in the previous excavation unit, all the structures were more or less rectilinear, and every room had its own four walls of about 30 cm in thickness. While most structures were only made up of one room, they are occasionally build wall to wall with doorways between them, leading to very thick walls. Not only the architecture, but also the materials we recovered from the second exposure were similar to what we had found before. Large amounts of iron slag in small pieces once again pointed to the intensive production of iron.

For the last week of fieldwork, we rented a car from Bamako to help us with the survey and take us and our materials back to the capital at the end of the week. Before this time, we had been walking the two kilometres to the site and back, transporting the equipment on the two motorbikes that belonged to members of the excavation team, and occasionally on donkey carts. But by the second week of March, the weather had heated up to around 45 °C, and we needed a better vehicle in order to be able to cover large amounts of ground effectively.

We started the survey being led by the griot of a local village, Abdoulaye Gajaga. A griot is a specialist in oral tradition, and so he was able to show us the sites that feature in local oral histories and explain the meanings attached to them. We then used another informant, the director of a nearby village museum, who has an interest in the areas history, and has been collecting oral traditions and archaeological materials found in the area around Douentza, and who was able to show us the sites he knew about. After this, we set out driving the car in straight transects with the help of a compass, while all except the driver and navigator stood on the back of the pick-up to scan the surrounding landscape for sites, which are easily identified as they are either easily visible mounds in an otherwise flat landscape, and/or covered in potsherds and laterite pebbles, resulting in distinctive reddish-brown patches.

In these ways we were able to record 14 sites, of which the majority were settlement mounds, but three of them were sites used for the smelting of iron. Judging from the pottery found on the surfaces of these sites, they were all more or less contemporaneous with the occupation of Tongo Maaré Diabal.  The pattern of settlement resulting from this survey is very different to those known from later times. In historical times, the population of this area were menaced by raiding for tribute or slaves. This means that only those who either submitted to the military forces or those under their protection were able to live in the plains. The others moved to defensive locations on the escarpments and plateaux. A systematic survey of these upland areas has not been undertaken. From the few sites that we do know of, however, it seems that intensive settlement of these areas did not start until the beginning of the second millennium AD – the time of abandonment of Tongo Maaré Diabal. Could it have been that a shift in political conditions at this time made the existing settlement pattern untenable?

The second unusual aspect to the survey results is the high level of evidence for ironworking. Almost every site we encountered worked or smelted iron, or both. In historical and ethnographic scenarios from Mali and the surrounding countries, being a blacksmith is a casted occupation, meaning that in order to become a blacksmith, you need to be born as one. Most villages have one or two active blacksmith’s workshops, and sometimes, blacksmith families set up their own wards outside villages or towns. To have a whole area full of settlements that are all producing and processing large quantities of iron, however, is very unusual. At the moment, our interpretation for this phenomenon is that the iron might have been traded to the Inland Niger Delta. This area, in which enormous cities conducted trade across the Sahara, and with other sub-saharan African regions, has no sources of iron ore, but had a large appetite for iron products, as excavations there have shown. These urban centres would have had to rely entirely on imports to satisfy their needs for agricultural implements and weapons of war. Could the sites we found during survey have been set up by specialised iron producers looking to exploit these needs?

We will need to wait for the completion of all analytical steps in order to draw deeper conclusions about the internal organisation of settlement at Tongo Maaré Diabal. But on the strength of the techniques employed in pottery-making, we can come to the conclusion that at least part of the population of the site has close links to the later and even present-day occupants of the area. While the drastic population replacements alluded to in the area's oral histories seem more and more unlikely, there do appear to be major ruptures in the Douentza region in the time leading up to the Tellem period. These changes transformed an area that was densely settled for around 800 years, with probable links to the economic and political powers of the day, into a depopulated area where the remaining inhabitants were forced to hide from raiding expeditions and conquering states.